AlphaGo and Deep Blue (which beat Garry Kasparov in 1997) cannot adapt to a small change in the rules of their game.

True if the documented design of AlphaGo (2015–2017 versions) and of IBM's Deep Blue (which beat Kasparov in the 1997 match) shows that neither could handle a small change to the rules of its game (for example, a different board size or a changed rule) without retraining or reprogramming by humans; and Deep Blue did beat Kasparov in 1997. Mostly true if that holds but a minor detail is off. Mostly false if one of the systems could in fact adapt to small rule changes without retraining. False if both could, or the Kasparov fact is wrong.

Set before the research began.

1. E1 Mastering the game of Go with deep neural networks and tree search - Nature Nature · 27 January 2016

   “These deep neural networks are trained by a novel combination of supervised learning from human expert games, and reinforcement learning from games of self-play.”

2. E2 Deep Blue | IBM IBM · 10 November 2025

   “In May of that year, it became the first computer system to defeat a reigning world chess champion in a match under standard tournament controls.”

3. E3 Deep Blue Artificial Intelligence journal (Elsevier), authors Murray Campbell, A. Joseph Hoane Jr., Feng-hsiung Hsu · 2002

   “The opening book in Deep Blue was created by hand, primarily by Grandmaster Joel Benjamin, with assistance from Grandmasters Nick De Firmian, John Fedorowicz, and Miguel Illescas.”

4. E4 AlphaGo - Wikipedia Wikipedia

   “AlphaGo and its successors use a Monte Carlo tree search algorithm to find its moves based on knowledge previously acquired by machine learning, specifically by an artificial neural network (a deep learning method) by extensive training, both from human and computer play.”

5. E5 Mastering the game of Go without human knowledge - Nature Nature (Springer Nature) · 19 October 2017

   “Here we introduce an algorithm based solely on reinforcement learning, without human data, guidance or domain knowledge beyond game rules.”

The AlphaGo papers (E1, E5) and the encyclopedia summary (E4) describe how its networks were trained and searched. None of them says whether AlphaGo could handle a changed board size or rule without being retrained. The Deep Blue paper (E3) shows that its opening book was built by hand, but it does not discuss rule changes either. The IBM page (E2) says Deep Blue beat a reigning world champion "in May of that year", but the quote names neither Kasparov nor the year, so even that part of the claim is not established from the quotes. The evidence does not settle the criteria either way.